Phase 2: Defining the analysis data model
IIW uses the term analysis data model to refer to the business model. The data warehousing development method (DWDM) delivery approach foresees the creation of the analysis data model before the creation of the logical and physical models. The analysis data model is business oriented. It is independent of any design or architecture. Defining the analysis data model is one of the most resource-intensive phases of the design process.
The analysis data model is a data model that specifies the normalized data structures required to represent the concepts defined in the conceptual model. As an analysis model, the analysis data model does not add any new business concepts to the content defined in the conceptual model. .
The analysis data model consists of 23 business areas, including party, place, claim, event, and so on. The analysis data model has over 1100 entities. There is a type hierarchy for each core entity, such as the type hierarchy for Claim core entity, as shown in Figure 8.
Figure 8. Claim type hierarchy
The supertype and its subtypes (known as natures) are connected using the parent-child relationship. The attributes are held in the appropriate level of the hierarchy.
The entities and attributes of the analysis data model are described with Enterprise Model Extender (EME)-specific properties. Additional Classification and Mappings tabs are shown in Figure 9.
Figure 9. EME-specific properties
Using the Classification option, you can verify the appropriate type of entities and attributes. The most common types are semantic entity for entities and semantic attribute for attributes. All attributes of the analysis data model are mapped to the conceptual model, as shown in Figure 10.
Figure 10. Mapping to the conceptual model
You can verify and analyze the model to identify the erroneous modeling or gaps. Model analysis is defined as the set of rules that can be checked against the analysis data model you created. For example, there is a rule to check for the existence of mappings for all attributes defined in the model, as shown in Figure 11.
Figure 11. Model analysis rules
If some mappings are missing, as shown in Figure 12, the appropriate error message appears after you execute the model analysis.
Figure 12. Error message sample
This section described how to define the analysis data model (business model), which is one of the most important tasks in the DWDM method. The person assigned to this activity needs both business knowledge and data modeling skills. The analysis data model is the basis for the definition of the logical data model described in Phase 3.